Highly drug-resistant tuberculosis (TB), including multidrug-resistant (MDR) and extensively drug- resistant (XDR) TB (collectively termed M/XDR-TB) represents a serious global threat to individual health, public health, and TB control, particularly in settings of high HIV prevalence. Yet, little is known about the epidemiology of M/XDR-TB, or optimal strategies for its detection, surveillance, clinical management, and prevention. Gold miners in South Africa have epidemic HIV infection and a high risk for all forms of TB. The overall goals of this proposal are to characterize M/XDR-TB in the study population of gold miners in South Africa, and to determine optimal strategies for detection of XDR-TB. Our objectives are to determine M/XDR-TB incidence, prevalence among all TB cases, risk factors, transmission patterns, and clinical outcomes in this population. In addition, we will evaluate laboratory performance characteristics of three rapid diagnostic tests for detection of M/XDR-TB in individual patients, determine and compare costs and feasibility of these tests, and use mathematical modeling to estimate the impact of various strategies for detection of XDR-TB at the population level. The study population has an HIV prevalence of approximately 24-30%, and TB case notification rates exceed 4000 per 100,000, almost 10-fold greater than for the rest of South Africa. Therefore, the proposed study population is an important sentinel population for XDR-TB in sub-Saharan Africa. The proposed studies will be performed using the enabling framework of the Thibela TB Project of the Consortium to Respond Effectively to the AIDS/TB Epidemic (CREATE). The study setting will be the 20 gold mine shafts with the associated population of ~70,000 employees that comprise the Thibela TB project. Application goals will be met through performance of a series of inter-related studies linked to a program of prospective laboratory surveillance for drug-resistant TB. The program of prospective laboratory surveillance will serve to identify and characterize (by drug-susceptibility status) all TB patients in the overall study population. This group of TB patients will serve as the study population for identification of clinical and demographic risk factors, outcomes, and transmission patterns of M/XDR-TB. In addition, the program of laboratory surveillance will incorporate an evaluation of new rapid tests for detection of rifampin resistance, and provide data relevant to future development of molecular tests for rapid detection of fluoroquinolone resistance. Information from this research project will be highly important from perspectives of individual and public health, and is directly responsive to research priorities identified by the WHO Global Task Force on XDR-TB. Information from this project will be invaluable for guiding strategies for prevention, diagnosis, and clinical management of M/XDR-TB throughout sub-Saharan Africa and other areas of high TB/HIV prevalence. PUBLIC HEALTH AND RELEVANCE: This project is relevant to public health because information from the project will guide development of strategies for prevention, diagnosis, and clinical management of M/XDR-TB patients throughout sub-Saharan Africa and other areas of high TB/HIV prevalence. This project is directly responsive to research priorities identified by the World Health Organization Global Task Force on XDR-TB. Specifically, identification of M/XDR-TB risk factors and transmission patterns will guide development of prevention strategies; rigorous assessment of rifampin resistance tests will guide development and implementation of optimal diagnostic strategies; and identification of factors associated with clinical outcomes will inform strategies for clinical management of M/XDR-TB patients.